Clock synchronization algorithm (physical)

For such a famous computer science article, I decided to read Lamport "Time", "Clock" and "Ordering events in a distributed system" . In addition to the frequently cited notes on logical clocks, it also offers an algorithm for synchronizing the physical clock (starting on page 562). I think I realized that this algorithm, however, on page 563, it lists two implementation rules that every process must follow. I just can't figure out what IR1 means:

IR 1 '. For each i, if Pi does not receive a message at physical time t, then Ci is differentiable for t and dCi (t) / dt> 0.

What does he mean by "Qi is differentiable"? Differentiated from what? And why should dCi (t) / dt be greater than 0? Are we not trying to make it as close to zero as possible?

This is perhaps a very stupid question that comes up with a really obvious answer, but at the moment I cannot figure it out. So help is greatly appreciated ...

+3


source to share


1 answer


This suggests that it Ci(t)

is a mathematically differentiable function of t

; those. derivative of Ci(t)

exists in t

. (AKA value Ci(t)

changes by some measurable amount when t

)

dCi(t)/dt > 0

simply means that the derivative of is Ci(t)

greater than zero. (AKA The rate at which the Ci(t)

change is positive at t

)



It's just defined Ci(t)

as a function that increments every time it Pi

doesn't receive a message.

+1


source







All Articles